Study Shows BPE Quantification on MRI Helps Predict Risk of Breast Cancer Recurrence

The combination of quantitative background parenchymal enhancement (BPE) measurements and tumor radiomics may help clinicians determine low, intermediate, and high risk of breast cancer recurrence.

In a retrospective study, recently published in Radiology, researchers utilized a computer algorithm to determine quantitative BPE measurements from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) exams of 127 women (mean age of 58) with diagnosed breast cancer. Utilizing the Oncotype DX recurrence score, they noted that 94 women in the cohort had low or immediate recurrence risk in comparison to 33 women with high recurrence risk.

The study authors found that multiple BPE measurements in ipsilateral and contralateral breasts were associated with increased risks, ranging from 27 to 66 percent higher risks, for breast cancer recurrence.

Here one can see dynamic, contrast-enhanced MRI scans for three women who were diagnosed with invasive ductal carcinoma. Based on the quantification of background parenchymal enhancement (BPE) over fibroglandular tissue, patient A was deemed to have a low risk of recurrence whereas patients B and C had intermediate and high recurrence risks respectively. (Images courtesy of Radiology.)

Some of the specific BPE measurements were as follows:

• 31 percent increased recurrence risk for relative absolute BPE within the fibroglandular tumor (FGT) to FGT volume (%BPEF/F) at a 20 percent cutoff rate in the ipsilateral breast

• 59 percent increased recurrence risk for relative BPEF to breast volume (%BPEF/B) at a 30 percent cutoff rate in the ipsilateral breast; and

• 41 percent increased recurrence risk for relative %BPEF/B at a 30 percent cutoff rate in the contralateral breast

“This retrospective study showed ipsilateral and contralateral DCE MRI BPE measures quantified in patients with breast cancer can be used to distinguish patients with high risk from those with low or intermediate recurrence … ,” wrote lead study author Dooman Arefan, Ph.D., an assistant professor of machine learning/AI, medical imaging and radiology at the University of Pittsburgh School of Medicine, and colleagues.

The researchers also found that combining DCE MRI BPE measurements with tumor radiomics had a more pronounced prognostic impact in estimating patient risk for distal recurrence.

“In the development set, we further showed that combining radiomic features with quantitative BPE measures in both breasts provides additional information beyond contralateral BPE measures (AUC, 0.94 vs. 0.73) … in informing distant recurrence risk,” noted Arefan and colleagues.

The researchers added that four key features utilized in the tumor radiomics included average wash-in slope, average peak enhancement, standard deviation of tumor pixel intensity and median tumor pixel intensity.

(Editor’s note: For related content, see “Study: Abbreviated MRI and DBT Offer Comparable Breast Cancer Detection in Dense Breasts,” “Can DWI MRI Offer a Viable Non-Contrast Alternative for Breast Cancer Assessment?” and “FDA Clears MRI-Compatible Robotics System for Breast Biopsy.”)

Beyond the inherent limitations of a single-center retrospective study, the researchers noted the cohort was restricted to women with estrogen receptor-positive breast cancer. The study authors also acknowledged that model performance in the study may have been affected by the use of systemic chemotherapy for a significant number of high-risk patients in the cohort.